Tree-Based Regime Switching in Yield-Curve Models

Summary

The study adapts the dynamic Nelson–Siegel framework, which represents the shape and movement of interest-rate curves, to allow its behavior to change across macroeconomic conditions. A tree-growing procedure divides observations into regimes by evaluating how well candidate macroeconomic splits explain the model’s marginal likelihood. This gives the resulting groups an economic interpretation tied to observable conditions and is presented as computationally simpler than conventional Markov-switching approaches.

Applied to U.S. Treasury yields, the analysis identifies meaningful changes in yield-curve regimes. It also finds that macroeconomic variables help predict the curve in a high federal funds rate regime, while comparable predictive value is not found in the other regimes. The result suggests that the connection between macroeconomic information and yields depends on the prevailing rate environment. The evidence is specific to the reported U.S. Treasury application; the excerpt provides no sample details, forecast metrics, or broader robustness results, so it does not establish that the same regime structure or predictive relationship holds in other markets or periods.

Key ideas

• A tree partitions macroeconomic conditions using the marginal likelihood of a dynamic Nelson–Siegel yield-curve model.
• The resulting regimes link changes in yield-curve behavior to observable economic variables.
• The method is described as easier to compute and interpret economically than Markov-switching models.
• In the U.S. Treasury application, macroeconomic predictors help explain yields when the federal funds rate is high.
• Predictive usefulness varies by regime, limiting claims that macroeconomic data consistently span the yield curve.

Tags

# Machine Learning and the Yield Curve: Tree-Based Macroeconomic Regime Switching


# Machine Learning and the Yield Curve: Tree-Based Macroeconomic Regime Switching









We explore tree-based macroeconomic regime-switching in the context of the dynamic Nelson-Siegel (DNS) yield-curve model. In particular, we customize the tree-growing algorithm to partition macroeconomic variables based on the DNS model's marginal likelihood, thereby identifying regime-shifting patterns in the yield curve. Compared to traditional Markov-switching models, our model offers clear economic interpretation via macroeconomic linkages and ensures computational simplicity. In an empirical application to U.S. Treasury yields, we find (1) important yield-curve regime switching, and (2) evidence that macroeconomic variables have predictive power for the yield curve when the federal funds rate is high, but not in other regimes, thereby refining the notion of yield curve ''macro-spanning''.

Shown in full with attribution under the source's licence. Licence: abstract CC0

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.